【大数据入门核心技术-Iceberg】（七）Iceberg与Flink整合

一、下载部署

wget https://search.maven.org/remotecontent?filepath=org/apache/iceberg/iceberg-flink-runtime-1.16/1.1.0/iceberg-flink-runtime-1.16-1.1.0.jar

cp iceberg-flink-runtime-1.16-1.1.0.jar /data/bigdata/flink-1.13.2/lib/

二、启动sql-client

1）修改flink-conf.yaml配置
vim $FLINK_HOME/conf/flink-conf.yaml

classloader.check-leaked-classloader: false
taskmanager.numberOfTaskSlots: 4

state.backend: rocksdb
execution.checkpointing.interval: 30000
state.checkpoints.dir: hdfs://hadoop101:9000/ckps
state.backend.incremental: true

2）cluster模式
（1）修改workers
vim $FLINK_HOME/conf/workers
#表示：会在本地启动3个TaskManager的 local集群
hadoop101
hadoop102
hadoop103

（2）启动Flink
$FLINK_HOME/bin/s